The proposed research is aimed at developing a computational tool which would reliably predict blood vessel remodeling resulting from vascular interventions. In planning interventions which result in flow alterations, clinicians often rely on intuition rather than solid scientific evidence. Reducing this uncertainty provides an exciting opportunity for computational modeling methods which could be used to explore various interventional options. Recent advances in patient-specific computational fluid dynamics (CFD) modeling indicate that these methods might now be sufficiently mature for this task. However, an important challenge for the adoption of postoperative flow modeling is the scarcity of well-controlled cases where accurate predictions of subsequent vascular changes have been demonstrated. Furthermore, CFD methods generally lack information about the flow in the proximal and distal circulation. We propose a novel approach where these flow boundary conditions will be obtained with in vivo measurements using time-resolved phase-contrast MR velocimetry (4D MRV). The proposed image-based CFD methodology will be applied on a patient-specific basis to three types of vascular interventions with differing functional and anatomic complexities. These include: maturation of arteriovenous fistulas created for hemodialysis access; fusiform cerebral aneurysms treated by occlusion of one or more proximal vessels; and finally, fusiform cerebral aneurysms treated by flow diverter stents. Currently, there is a high incidence of unsuccessful treatment outcomes in both fusiform aneurysms and arteriovenous fistulas. The proposed image-based CFD methodology can evaluate postoperative values of relevant hemodynamic descriptors and thus identify early indicators of a likely fistula failure, or flag as unsuitable, treatments of fusiform aneurysms that could lead to negative developments such as thrombotic occlusion of a vital perforator. It is expected that this could help in selecting appropriate treatment options and thus increase the number of favorable outcomes. UCSF/VASF has international leaders in vascular surgery, radiology and biomedical research. The full array of clinical and research facilities at UCSF/VA will be available for the proposed research studies. The team assembled to work on this project has a long history of successful and productive collaboration. The vascular/neurovascular surgeons will identify candidate subjects from patients scheduled for treatment by one of the interventional procedures specified above. The CFD-predicted vessel adaptations will be correlated to in vivo observations in order to fine-tune and validate our modeling methods. Once the efficacy and limitations of this methodology are established, it can be used for prospective patient-specific modeling of vascular interventions in order to provide guidance to vascular and neurovascular surgeons. Successful completion of the project will lead to a modeling tool capable of predicting a priori the impact of various treatment options on postoperative vessel remodeling, thereby permitting stratification of patients and individualized treatment.